TECHNICAL REPORT. External quality assurance scheme for typing of verocytotoxin-producing E.coli (VTEC)

TECHNICAL REPORT External quality assurance scheme for typing of verocytotoxin-producing E.coli (VTEC) www.ecdc.europa.eu

ECDC TECHNICAL REPORT External quality assurance scheme for typing of verocytotoxin-producing E. coli (VTEC) as part of the European Food- and Waterborne Diseases and Zoonoses Programme

This report was commissioned by the European Centre for Disease Prevention and Control (ECDC), coordinated by Taina Niskanen and Therese Westrell, and produced by Statens Serum Institut (Denmark) on behalf of the European Food- and Waterborne Diseases and Zoonoses Programme. Author Flemming Scheutz, WHO Collaborating Centre for Reference and Research on Escherichia and Klebsiella, Statens Serum Institut (SSI). Suggested citation: European Centre for Disease Prevention and Control. External quality assurance scheme for typing of verocytotoxin-producing E. coli (VTEC). Stockholm: ECDC;2012. Stockholm, April 2012 ISBN 978-92-9193-340-2 doi 10.2900/32337 European Centre for Disease Prevention and Control, 2012 Reproduction is authorised, provided the source is acknowledged.

TECHNICAL REPORT EQA for typing of VTEC Contents Abbreviations... iv Executive summary... 1 1 Background... 3 2 Introduction... 4 3 Materials and methods... 5 3.1 Organisation... 5 3.2 Selection of strains... 5 3.3 Carriage of strains... 5 3.4 Testing... 5 3.5 Data analysis... 5 4 Results... 8 4.1 Serotyping... 8 4.2 Phenotypic detection... 8 4.3 Genotypic detection... 8 5 Discussion... 10 6 Conclusion... 12 References... 13 Annex 1. Results of the participants by strain... 14 Annex 2. List of the participants... 18 ECDC-funded EU/EEA participants... 18 ECDC-funded non-eu/eea country... 18 Self-funded participants... 18 Annex 3. Prevalence of sero- and virulence types (Enter-net)... 20 Annex 4. Individual results for all laboratories... 21 iii

EQA for typing of VTEC TECHNICAL REPORT Abbreviations DNA ECDC EEA EIA EPEC EQA EU FWD-Net GFN HUS NSF PCR PFGE QMS SF SF SSI VCA VTEC WHO Deoxyribonucleic acid European Centre for Disease Prevention and Control European Economic Area Enzyme immunoassay Enteropathogenic Escherichia coli External quality assurance European Union Food- and Waterborne Diseases Network WHO Global Foodborne Infections Network Haemolytic uremic syndrome Non-sorbitol-fermenters Polymerase chain reaction Pulse field gel electrophoresis Quality management systems Self-funded Sorbitol-fermenters Statens Serum Institut, Denmark Vero cell assay Verocytotoxin-producing Escherichia coli World Health Organization iv

TECHNICAL REPORT EQA for typing of VTEC Executive summary Main findings Forty-five laboratories from 38 countries participated in the second international external quality assurance scheme by ECDC on typing of VTEC. Twenty-six of the laboratories were from 23 EU/EEA countries and nineteen laboratories were from 15 non-eu countries. An average of 78% of the EU/EEA countries were able to correctly O:H serotype the ten VTEC strains that were included in the EQA. Correct phenotypic characterisation was generally very high over 92% in the EU/EEA laboratories. About 86% of the non-eu/self-funded (SF) laboratories were able to correctly O:H serotype the ten VTEC strains. Correct phenotypic characterisation was higher than 85% in the non-eu/sf laboratories. Accurate gene detection was very high except for subtypes vtx1d and vtx2f. Accurate detection of vtx2c was at 100% for EU/EEA laboratories. In non-eu/sf laboratories, correct detection of vtx2c was 95%. Improvements need to be made in some laboratories regarding the methodology used for specific subtypes, preparation of template DNA to avoid cross-contamination and choice of primers for the detection of the eae gene. The fact that only 9/17 (53%) laboratories detected verocytotoxin production in the O157:H7 strain underlines the importance of genotypic detection. This report presents the results of the second round of the external quality assurance (EQA) scheme for typing of verocytotoxin-producing Escherichia coli (VTEC) funded by the European Centre for Disease Prevention and Control (ECDC). The EQA was conducted between December 2009 and June 2010 and included O:H serotyping, detection and typing of eae, vtx1, vtx2 and ehxa genes, phenotypic detection of verocytotoxin/shiga toxin-production, fermentation of sorbitol and production of β-glucuronidase and enterohaemolysin for VTEC. In February 2010, the laboratories of the Food- and Waterborne Diseases Network (FWD-Net) were contacted and asked which parts of the scheme they wanted to participate in (serotyping, virulence typing or both). In March 2010, the WHO Global Foodborne Infections Network (GFN) kindly offered to translate the invitation to participate in the EQA into Chinese, French, Spanish, Portuguese and Russian, and circulated the invitation through the GFN network in April and May of 2010. Forty-five laboratories from 38 countries participated, of which 23 were from the EU/EEA countries: Austria, Belgium, Bulgaria, Cyprus, the Czech Republic, Denmark, England, Finland, France (two laboratories), Germany (three laboratories), Greece, Hungary, Ireland, Italy, Lithuania, Luxembourg, the Netherlands, Norway, Poland, Romania, Slovenia, Spain and Sweden. Only one EU/EEA laboratory per country was funded by the ECDC EQA programme. In Denmark, nine regional hospitals also participated in this round of EQA and in addition to the ten VTEC strains the Danish EQA programme also included typing of five non-vtec strains. Malta, Portugal and Scotland accepted the invitation to participate but did not meet the extended deadline of 31 August 2010 for submission of results. Eighteen self-funded (SF) laboratories from 14 non-eu countries participated: Argentina, Australia, Bangladesh, Brazil, Canada (three laboratories), India, Japan, Mexico, New Zealand, Philippines, South Africa, Switzerland, USA (three laboratories) and Vietnam. In addition, one EU candidate country (Turkey) was funded by ECDC. The participating laboratories could choose whether to participate in full typing or only a selection of the methods and/or strains. Participation for full O:H serotyping among the EU/EEA countries was 57 70%, which was slightly lower than participation by the non-eu/sf laboratories. Of these, an average of 78% of EU/EEA laboratories could correctly O:H serotype all the strains. The more common a serotype is, the higher the quality of typing was, ranging from 100% correct typing of O157:H7 to 38% of O41:H26 among the EU/EEA countries and 71% of O128ac:[H2] and O177:[H25] among the non-eu/sf laboratories. Participation for EU/EEA countries in phenotypic detection was 57% for β-glucuronidase, 39% for haemolysin, 35% for verocytotoxin while fermentation of sorbitol was 83%. The percentages of results that were correct in EU/EEA countries were 92% for β-glucuronidase, 98% for haemolysin and 96% for verocytotoxin while fermentation of sorbitol was 97%. Genotypic detection of eae, vtx1 and vtx2 was performed by 21/23 (91%) of the EU/EEA participants with 95% correct detection, followed by the detection of the ehxa gene which was detected 99% correctly by 87% of the EU/EEA participants. Surprisingly, only 5/8 (63%) of the EU/EEA laboratories were able to phenotypically detect verocytotoxin in the O157:H7 strain (JJ10) which encodes subtype vtx2c a subtype known to be weakly expressed due to a weak 1

EQA for typing of VTEC TECHNICAL REPORT upstream promoter. None of the EU/EEA laboratories, however, failed to detect the gene for vtx2. Detection of the gene of vtx2f in strain CC3 was correctly made by 15/21 (71%) of the EU/EEA laboratories. Detection of vtx1d was correct in 20/21 (95%) EU/EEA laboratories. False positive vtx1 genes were reported once by one EU/EEA laboratory. One EU/EEA laboratory detected three false positive vtx2 strains. The eae gene was detected correctly by all EU/EEA laboratories except for one negative. Participation in full O:H serotyping among the non-eu/sf laboratories was 64 82%. An average of 86% non-eu/sf laboratories could correctly O:H serotype all the strains. Participation of non-eu/sf countries in phenotypic detection was 55% for β-glucuronidase, 50% for haemolysin, and 41% for verocytotoxin while fermentation of sorbitol was 77%. The percentages of results that were correct in non-eu/sf countries were 87% for β- glucuronidase, 85% for haemolysin, and 87% for verocytotoxin while fermentation of sorbitol was 96%. Genotypic detection of eae, vtx1 and vtx2 was performed by 20/22 (91%) of the non-eu/sf participants with results of 92% correct detection or above, followed by the detection of the ehxa gene which was detected 98% correctly by 73% of the non-eu/sf participants. Only 4/9 (44%) of the non-eu/sf laboratories were able to phenotypically detect verocytotoxin in the O157:H7 strain (JJ10) which encodes subtype vtx2c. Only one of the non-eu/sf laboratories, however, failed to detect the gene for vtx2. Detection of the gene of vtx2f in strain CC3 was correctly made by 7/20 (35%) of non-eu/sf laboratories. Detection of vtx1d was correct in 13/20 (65%) non-eu/sf laboratories. False positive vtx1 genes were reported twice by one non-eu/sf laboratory. One non-eu/sf laboratory in two strains and another laboratory in one strain detected three false positive vtx2 strains. The non-eu/sf laboratories reported four false positive eae genes, one failed to detect one eae gene and one failed to detect four eae genes. 2

TECHNICAL REPORT EQA for typing of VTEC 1 Background The European Centre for Disease Prevention and Control (ECDC) is a European Union (EU) agency with a mandate to operate the dedicated surveillance networks and to identify, assess, and communicate current and emerging threats to human health from communicable diseases. Within its mission, ECDC shall foster the development of sufficient capacity within the Community for the diagnosis, detection, identification and characterisation of infectious agents which may threaten public health. The Centre shall maintain and extend such cooperation and support the implementation of quality assurance schemes. (Article 5.3, EC 851/2004 1 ). External quality assurance (EQA) is part of quality management systems (QMS) and evaluates performance of laboratories, by an outside agency on material that is supplied especially for the purpose. ECDC s disease specific networks organise a series of EQA schemes for EU/EEA countries. In some specific networks non-eu/eea countries are also involved in the EQA activities organised by ECDC. The aim of EQA is to identify needs for improvement in 2 laboratory diagnostic capacities relevant to surveillance of the diseases listed in Decision No 2119/98/EC and to ensure comparability of results in laboratories from all EU/EEA countries. The main purposes of EQA schemes include: assessment of the general standard of performance; assessment of the effects of analytical procedures (method principle, instruments, reagents, calibration); evaluation of individual laboratory performance; identification and justification of problem areas; providing continuing education; identification of needs for training activities. For many years the European Commission funded Enter-net, an international surveillance network for national reference laboratories and surveillance centres on selected human gastrointestinal infections. Custodianship of Enter-net migrated to ECDC in October of 2007. In 2008, a framework contract for external quality assurance for Salmonella and verocytotoxin-producing E. coli (VTEC) was put in place for the years 2008 2011. The contract for the VTEC EQA was awarded to the laboratory of the International Escherichia and Klebsiella Centre (WHO) hereafter referred to as The Centre at Statens Serum Institut (SSI) in Denmark. This laboratory now arranges annual EQA rounds for the national reference laboratories in the EU and EEA countries on serotyping and virulence typing for VTEC. 1 Regulation (EC) no 851/2004 of the European Parliament and of the Council of 21 April 2004 establishing a European Centre for Disease Prevention and Control. OJ L 142, 30.4.2004, p. 1 11. 2 Decision No 2119/98/EC of the European Parliament and of the Council of 24 September 1998 setting up a network for the epidemiological surveillance and control of communicable diseases in the Community. OJ L 268, 3.10.1998, p. 1 7. 3

EQA for typing of VTEC TECHNICAL REPORT 2 Introduction The International Escherichia and Klebsiella Centre at the Statens Serum Institute (SSI) in Denmark has played a leading role in establishing a worldwide international network of quality evaluation and assurance for the typing of E. coli since 2002. The first two ring trials with serotyping were organised in 2002 and 2003. In 2005, the third ring trial was launched, including serotyping, virulence typing and typing by pulse field gel electrophoresis (PFGE) of E. coli. The PFGE part of the ring trial was co-ordinated with PulseNet Europe 3 as part of the PulseNet Europe Feasibility Study of VTEC. The fourth international ring trial in 2006 was, apart from the regular O:H serotyping, centred on the capacity to detect the vtx genes. In 2007, the fifth ring trial EQA scheme for sero- and virulencetyping of VTEC was arranged and in 2008, the Centre was awarded the tender for EQA for VTEC organised by ECDC. State-of-the-art characterisation of VTEC includes O:H serotyping in combination with a few selected virulence genes, i.e. the two genes for verocytotoxin VT1 (vtx1) and VT2 (vtx2), and the eae gene associated with the attaching and effacing lesion of enterocytes also seen in enteropathogenic E. coli (EPEC). The combination of some of these toxin determinants is clinically relevant in that some subtypes of VT2 (VT2a in eae-positive VTEC and two activatable VT2d variants in eae-negative VTEC) seem to be highly associated with the serious sequelae haemolytic uremic syndrome (HUS) [1,2], but VT2c-positive VTEC can also cause HUS [2,3]. Other specific subtypes or variants of VT1 and VT2 are primarily associated with milder course of disease [2,3], and VT2e-positive VTEC strains are probably not pathogenic to humans [4]. Our understanding of the epidemiology of the VT subtypes is therefore important for reducing the risk of VTEC infection and for the surveillance of VTEC. Finally, some of the existing subtyping methods using a combination of specific polymerase chain reaction (PCR) and restriction fragment length polymorphism (RFLP) are inadequate and may result in misleading conclusions. For example, typing of vtx2 has been based on the absence of the PstI site as an indicator of the presence of the mucusactivatable stx2d subtype [1,5 7]. However, the PstI site is also absent in six variants of vtx2a, in two variants of vtx2c, in stx2f and in all four variants of subtype stx2g. Furthermore, the most commonly detected VTEC serotype O157:H7 may be divided into two groups: one with the unusual property of failing to ferment sorbitol within the first 20 hours of incubation (the non-sorbitolfermenters, NSF) and one highly virulent variant of O157 which will ferment sorbitol (SF). NSF O157 is most often also characterised by failure to produce β-glucuronidase. Approximately 75% of all VTEC also produce enterohaemolysin, a toxin which can cause lysis of erythrocytes. Enterohaemolysin may either be detected phenotypically on sheep blood agar plates or by detection of the gene ehxa encoding enterohaemolysin. Proficiency testing therefore included O:H serotyping, detection and typing of eae, vtx1, vtx2 and ehxa genes, phenotypic detection of Verocytotoxin/Shiga toxin-production through vero cell assay (VCA) or enzyme immunoassay (EIA), fermentation of sorbitol, and production of β-glucuronidase and enterohaemolysin. 3 PulseNet Europe is the molecular surveillance network for food-borne infections in Europe. http://www.pulsenetinternational.org/networks/pages/europe.aspx 4

TECHNICAL REPORT EQA for typing of VTEC 3 Materials and methods 3.1 Organisation The second round of the VTEC EQA funded by ECDC was arranged in December 2009 and included O:H serotyping, detection and typing of eae, vtx1, vtx2 and ehxa genes, phenotypic detection of verocytotoxin/shiga toxinproduction, fermentation of sorbitol and production of β-glucuronidase and enterohaemolysin for VTEC. Invitations were sent to the VTEC laboratory experts in the FWD network in February 2010. In March and April 2010, the invitation to participate was further circulated through the WHO Global Foodborne Infections Network (GFN), translated into Chinese, French, Spanish, Portuguese and Russian. Nine Danish regional hospitals also participated in this round. In total, 45 laboratories participated, of which 23 were in EU/EEA countries and one was a candidate country whose participation was also funded by ECDC (Annex 2). By the end of May 2010, the strains had been sent to the laboratories, which were required to return their results by 18 June. Each laboratory received their individual results in September 2010. Due to a delay in shipping the strains, the overall results could only be presented in brief at the second annual meeting of the European Food- and Waterborne Diseases and Zoonoses Network in Dublin, 22 23 June 2010. 3.2 Selection of strains Strains were selected for the EQA programme based on three criteria: a) they should represent the most commonly reported strains, b) they should remain stable during the preliminary testing period at the organising laboratory and c) they should be easy to type. Nine of the ten strains included in the programme were carefully chosen from the top thirty sero-and virulence types of 5 160 strains reported to the dedicated surveillance network Enter-net in the years 2000 2006 (see Annex 3). One of these strains, O157:H7 only encoding vtx2c, was chosen because this subtype of verocytotoxin is often poorly expressed leading to failure in phenotypic detection. Another strain, O128:[H2] encoding vtx2f, was chosen to represent this particular virulence profile in order to test the participants capacity to detect vtx2f. The tenth strain encoding vtx1d was added to the panel because vtx1d is rarely isolated from humans and inclusion of this strain could give an indication of whether this is due to a technical detection problem. The characteristics of the 10 VTEC strains included are listed in Table 1. 3.3 Carriage of strains By the end of May 2010, all strains were packed and mailed as dangerous goods to all but one EU (Malta) and one non-eu/sf (USA) participant. Some of the parcels were delayed in delivery due to holding at customs and courier-related problems, and therefore not all the involved laboratories were able to meet the deadline (18 June 2010) for returning their results. Later, strains were mailed as diagnostic specimens (UN3373) with a door-to-door courier to the participants. All parcels were received in good time to meet the extended dead line of 31 August. 3.4 Testing The EQA tests included O:H serotyping, detection and typing of eae, vtx1, vtx2 and ehxa genes, phenotypic detection of verocytotoxin/shiga toxin-production, fermentation of sorbitol and production of β-glucuronidase and enterohaemolysin for VTEC. Subtyping of vtx genes was also encouraged on a voluntary basis but suffers from the lack of universal typing procedures and commonly accepted nomenclature. Participants were requested to test for additional virulence genes at their own convenience and capacities. This voluntary and additional testing is not a core part of the EQA programme but is meant as a source for sharing information on the capacities found within the network of participating laboratories. It provides additional information on the test strains, which may be valuable if and when laboratories wish to set up new tests. 3.5 Data analysis When the results from each individual laboratory were received by the Centre, they were immediately entered into the EQA database and scores provided to the relevant laboratory. 5

EQA for typing of VTEC TECHNICAL REPORT Table 1. Characteristics of the 10 VTEC strains used in the second VTEC EQA 2009 2010 Ranking a Strain No. Serotype Sorbitol fermentation β-glucuronidase activity Haemolysin production VCA eae vtx1 vtx2 ehxa Subtyping of vtx genes b Other virulence genes Comments 19 AA1 O174:H 8 pos. pos. neg. pos. neg. pos. pos. neg. vtx1c + vtx2b 23 BB2 O55:H 7 neg. pos. neg. pos. pos. pos. neg. neg. vtx1a asta 9 c CC3 O128ac:[H2] d pos. pos. neg. pos. pos. neg. neg. pos. vtx2f bfpa, asta 16 DD4 O177:[H25] d neg. pos. Ent. pos. pos. neg. pos. neg. vtx2c + vtx2d Lactose neg. 7 EE5 O111:[H8] d pos. pos. pos. pos. pos. pos. pos. neg. vtx1a + vtx2a 12 FF6 O113:H4 pos. pos. Ent. pos. neg. pos. pos. neg. vtx1c + vtx2b asta 2 GG7 O103:H2 pos. pos. Ent. pos. pos. pos. neg. neg. vtx1a 3 HH8 O26:H11 pos. pos. neg. pos. pos. pos. neg. neg. vtx1a NR ii9 O41:H26 pos. pos. neg. pos. neg. pos. neg. neg. vtx1d 1 e JJ10 O157:H 7 neg. neg. Ent. (pos.) pos. neg. pos. neg. vtx2c asta All results from WHO Collaborating Centre for Reference and Research on Escherichia and Klebsiella according to standard of accreditation DS/EN ISO/IEC 17025 described in DANAK accreditation No. 397. ALL strains were negative for the following genes: EAF, saa, ipah, aata (CVD432), LT, STh and STp. a Sero-and virulence types reported to Enter-net in 2000 2006 are ranked by prevalence (highest to lowest) (see Annex 3). b Results were obtained using a new subtyping PCR protocol developed by the WHO Collaborating Centre for Reference and Research on Escherichia and Klebsiella for the subtyping of the vtx1 and vtx2 genes not included in the accredited analyses (In prep.). c A strain of O128:[H2] encoding vtx2f was chosen to represent this virulence profile in order to test participants capacity to detect vtx2f. d Non-motile strains where H types were confirmed by flic PCR/RFLP by five participating laboratories and not by the organising laboratory at the Centre. Submission of NM as a result was also considered correct. e A strain of O157:H7 only encoding vtx2c was chosen because this subtype of verocytotoxin is often poorly expressed leading to failure in phenotypic detection. Genes detected asta: pss126. Enteroaggregative heat stable toxin (EAST1). bfpa: MSD207. Bundle forming pilus gene probe. eae: CVD434. E. coli attaching and effacing gene probe. ehxa: CVD419. Plasmid encoded O157-enterohaemolysin. vtx1: NTP705. Verotoxin 1; Almost identical with the Shiga toxin. vtx2 : DEP28. Verotoxin 2; Variants exist. Approx. 60% homology to VT1. Genes not detected all strains were negative for the following genes: aata: Plasmid marker (formerly the CVD432) encoding a dispersin translocator in Enteroaggregative E. coli. EAF: JPN16. EPEC Adherence Factor gene probe. ipah: WR390. Invasion plasmid antigen. These genes are found in several copies chromosomally as well as on plasmids. LT (eltb): G119. Heat labile enterotoxin. Almost identical to cholera toxin. saa: STEC Auto agglutinating adhesion; PCR. ST h (STIb; estah): DAS100. Heat stable enterotoxin (human variant). ST p (STIa; esta): DAS101. Heat stable enterotoxin (porcine variant). 6

TECHNICAL REPORT EQA for typing of VTEC Table 2. Summary of participation rates * for each test EU/EEA laboratories funded by ECDC and non-eu/sf laboratories EU/EAA participants Non-EU/selffunded (SF) laboratories Serotyping Phenotypic detection Genotypic detection O:H Serotype O group H type Sorbitol fermentation 13 16/23 57 70% 14 18/22 64 82% 15 19/23 (19 for O26, O103 & O157) 65-83% 19 21/22 (21 for O26, O103, O111 and 22 for O157) 86 95% 13-16/23 (16 for H7 and H11) 57-70% 14-18/22 (17 for H7) 64-82% 19/23 83% 17/22 77% β-glucuronidase activity 13/23 57% 12/22 55% Haemolysin production 9/23 39% 11/22 50% VCA/EIA eae vtx1 vtx2 ehxa Subtyping of vtx 8/23 21/23 21/23 21/23 21/23 3 11/23 35% 9/22 41% 91% 20/22 91% 91% 20/22 91% 91% 20/22 91% 91% 16/22 73% 13 48% 4 10/22 18 45% * Number of laboratories participating per method. 7

EQA for typing of VTEC TECHNICAL REPORT 4 Results The overall results for both the EU/EEA and the non-eu/sf laboratories are listed in Table 2. The specific results for the EU/EEA laboratories are presented in Table 4 (Annex 1) and for the non-eu/sf laboratories in Table 5 (Annex 1). Subtyping results for vtx are listed in Table 6 (Annex 1) for both EU/EEA and non-eu/sf laboratories and results for additional virulence genes presented in Table 7 (Annex 1). 4.1 Serotyping Participation of EU/EEA laboratories in O:H serotyping was slightly lower than the non-eu/sf laboratories (13 16 laboratories (57 70%) and 14 18 laboratories (64 82%), respectively). An average of 78% (average 11.4; median 11) of EU/EEA laboratories could correctly O:H serotype the ten strains, compared with 86% (average 12.9; median 12.5) of non-eu/sf laboratories. Correct O:H serotyping ranged from 100% correct typing of O157:H7 for all participating laboratories to 38% of O41:H26 among the EU/EEA countries and 71% of O128ac:[H2] and O177:[H25] among the non-eu/sf laboratories. Among the EU/EEA participants, O157:H7 was correctly serotyped by all laboratories, followed by serotypes O55:H7 (94%), O103:H2 (93%), O26:H11 (88%), O128:H2 (79%), O111:[H8] (79%) and O113:H4 (79%) (Annex 1, Table 4). Lower scores were obtained for O174:H8 (71%) followed by O177:[H25] (57%) and O41:H26 (38%). The latter (strain ii9) was included because of the presence of vtx1d, see Annex 1, Table 6. Thus, the less common a serotype is, the more difficult it was for the participants to serotype the strain correctly. 4.2 Phenotypic detection Participation of EU/EEA and non-eu/sf countries, respectively, in phenotypic detection was 57% and 55% for β- glucuronidase, 39% and 50% for haemolysin, and 35% and 41% for verocytotoxin while fermentation of sorbitol was 83% and 77%. Of the results for EU/EEA and non-eu/sf countries, respectively, the percentage correct were 92% and 87% for β-glucuronidase, 98% and 85% for haemolysin, and 96% and 87% for verocytotoxin, while fermentation of sorbitol was 97% and 96%. Five out of eight (63%) EU/EEA and three out of nine (33%) non- EU/SF laboratories were able to phenotypically detect verocytotoxin in the O157:H7 strain (JJ10) which encodes subtype vtx2c. However, only one laboratory (non-eu/sf) failed to detect the gene for vtx2. 4.3 Genotypic detection 4.3.1 Detection of virulence genes eae, vtx1, vtx2 and ehxa There was a similar level of participation in gene detection of virulence genes eae, vtx1 and vtx2 for EU/EEA laboratories as non-eu/sf laboratories, being performed by 21/23 (91%) of the EU/EEA and 20/22 (91%) of the non-eu/sf participants, while genotypic detection of ehxa was performed by a higher percentage (87% or 20/23) of the EU/EEA laboratories than the non-eu/sf laboratories (73% or 16/22). All four genes were detected at a very high score of 95% to 99.5% for EU/EEA laboratories and 92% to 98% of the non-eu/sf participants (Annex 1, Tables 4 and 5). Failure to detect the gene of vtx2f in strain CC3 occurred in 6/21 (29%) of EU/EEA and in 13/20 (65%) of non-eu/sf laboratories. Detection of vtx1d only failed in 1/21 (5%) EU/EEA laboratories compared to 7/20 (35%) non-eu/sf laboratories. The eae gene was detected correctly by all EU/EEA laboratories except for one that failed to detect eae in strain number EE5 (serotype O111:[H8]). Among the non-eu/sf laboratories, two false positive eae genes were reported by one laboratory in strains AA1 (serotype O174:H8) and FF6 (serotype O113:H4) and two laboratories reported a single false positive eae detection: one in strain AA1 (serotype O174:H8) and one in ii9 (serotype O41:H26). One laboratory failed to detect the eae gene in DD4 (serotype O177:[H25]) and one laboratory failed detection of the eae gene in four strains: CC3 (serotype O128ac:[H2]), DD4 (serotype O177:[H25]), GG7 (serotype O103:H2) and HH8 (serotype O26:H11). One EU/EEA laboratory detected one false positive vtx1 gene in strain CC3 (serotype O128) and one laboratory detected three false positive vtx2 strains: BB2 (serotype O55:H7), GG7 (serotype O103:H2) and HH8 (serotype O26:H11). One non-eu/sf laboratory reported two false positive vtx1 genes in strains CC3 (serotype O128) and DD4 (serotype O177:[H25]) and another laboratory in strain in ii9 (serotype O41:H26). 4.3.2 Subtyping of vtx1 and vtx2 The number of laboratories participating in subtyping of vtx genes was generally low, ranging from 3 to 11 (13 48%) among EU/EEA laboratories and from 4 to 10 (32 45%) among non-eu/sf laboratories. Correct 8

TECHNICAL REPORT EQA for typing of VTEC nomenclature and subtyping of both vtx1 and vtx2 genes ranged from 2/6 to 3/3 corresponding to 30 100% in EU/EEA laboratories and from 3/9 to 7/7 corresponding to 33 100% for the non-eu/sf laboratories. Ten out of 11 EU/EEA laboratories correctly typed vtx2f and 4/5 typed vtx1d correctly. The numbers of non-eu/sf laboratories correctly typing vtx2f were 7/7 and for vtx1d, 3/4. Subtyping results for vtx genes are presented in Annex 1, Table 6. 4.3.3 Detection of other virulence genes Results for additional virulence genes are presented in Annex 1, Table 7. Subtyping of the eae gene was obtained for the seven eae-positive strains by 2 4 laboratories with similar results. Two to five LEE (Locus for enterocytes effacement) encoded genes were detected in these eae-positive strains, and the tccp gene was found by one laboratory in all these except for strain HH8 (O26:H11). One to three non-lee encoded genes (nle) were detected in six strains. Long polar fimbrial genes (lpf) using a non-comparable nomenclature were detected in nine of the strains by four laboratories. Of the four laboratories examining for iha, only three strains were found positive by all four laboratories, while three strains were found positive three times, one strain twice and one strain once. One laboratory found all ten strains negative for bfpa and another all ten strains negative for bfpb. One laboratory found six strains positive for bfpa including strain CC3 (serotype O128ac:[H2]), which was found positive by the Centre in two independent tests. Three to seven laboratories found strains BB2 (O55:H7), CC3 (O128ac:[H2]), FF6 (O113:H4) and JJ10 (O157:H7) positive for asta, efa1 (or the similar toxb) was found by 3 4 laboratories in strains BB2 (O55:H7), EE5 (O111:[H8]), GG7 (O103:H2), HH8 and JJ10 (O157:H7). 9

EQA for typing of VTEC TECHNICAL REPORT 5 Discussion The participation rate varied substantially between the different tests included in the EQA, being highest for the genotypic detection of eae and the vtx genes and lowest for the subtyping of vtx. The serotyping results ranged from 100% correct typing of O157:H7 for all laboratories to 38% correct for O41:H26 among the EU/EEA countries, and 71% for O128ac:[H2] and O177:[H25] among the non-eu/sf laboratories. The general trend corresponded to how commonly the serotypes have been detected in humans in recent years, as the more common a serotype is, the higher the performance was (Table 3). Among the EU/EEA participants, O157:H7 was correctly serotyped by all laboratories followed by serotypes O55:H7 (94%), O103:H2 (93%), O26:H11 (88%), O128:H2 (79%), O111:[H8] (79%) and O113:H4 (79%). Except for O103:H2 and O113:H4, these serotypes may also be EPEC serotypes, which have traditionally been looked for by many laboratories. Furthermore, commercial kits focus on these particular serotypes. Lower scores were obtained for the two strains with more recent O groups that lie outside the traditional panel with O174:H8 (71% correctly serotyped) followed by O177:[H25] (57%) and O41:H26 (38%), the latter included because of the presence of vtx1d. Participation in phenotypic detection was relatively low. The highest participation was for sorbitol fermentation with high average scores for both EU/EEA and non-eu/sf laboratories. The sorbitol fermentation test is important for identifying the highly virulent sorbitol-fermenting O157:H7 clone. Surprisingly, only 5/8 (63%) EU/EEA and 3/9 (33%) of the non-eu/sf laboratories were able to phenotypically detect verocytotoxin in the O157:H7 strain (JJ10) which encodes subtype vtx2c a subtype which has been shown to be weakly expressed due to a weak upstream promoter. However, only one laboratory (non-eu/sf) failed to detect the gene for vtx2.. Based on the results of vtx subtyping, it is recommended that six EU/EEA and 13 non-eu/sf laboratories review their genetic detection methodology for vtx2f and that one EU/EEA and seven non-eu/sf laboratories review their detection for vtx1d. Two EU/EEA and four non-eu/sf laboratories should review their sample preparation procedures of template DNA for PCR, bearing in mind that false positive contamination is best minimised when DNA template preparation is done separately from running the PCR reaction. One non-eu/sf laboratory should review their choice of primers for detection of the eae gene. Table 3. Serotyping scores by the EU/EEA and non-eu/sf participants in relation to the ranking of the serotype in Europe*. Table sorted in descending order of scores for EU/EEA laboratories. Ranking Strain No. Serotype Percentage of results correct for O:H serotyping by the EU/EEA laboratories n = 14 17 Percentage of results correct for O:H serotyping by the selffunded laboratories n = 14 18 Prevalence of reported O:H serotype in the EU/EEA 1 JJ10 O157:H7 100% 100% 28% 23 BB2 O55:H7 94% 100% 0.3% 2 GG7 O103:H2 93% 93% 11% 3 HH8 O26:H11 88% 89% 9% 9 CC3 O128ac:[H2] 79% 71% 2% 7 EE5 O111:[H8] 79% 79% 2% 12 FF6 O113:H4 79% 86% 1% 19 AA1 O174:H8 71% 93% 0.4% 16 DD4 O177:[H25] 57% 71% 1% NR ii9 O41:H26 38% 79% <1% Average score 78% 86% Cumulated prevalence 57% * Ranking based on the number of O:H serotyped cases reported to Enter-net in 2000 2006 (see also Annex 3, Table 8). n: Number of laboratories participating per method. NR: not ranked in the Enter-net database. Comments on discrepancies and validity of vtx subtyping and other virulence genes The nomenclature of the subtypes and variants of the vtx genes varies see footnote b in Table 6 (Annex 1) for how interpretation was implemented for this EQA. The previously proposed nomenclature is currently being finalised. An update on the revised proposal for the revision of the vtx nomenclature was presented on 30 October 2009 at the fourth annual workshop of the National Reference Laboratories for E. coli in the EU, organised by the 10

TECHNICAL REPORT EQA for typing of VTEC Community Reference Laboratory for E. coli, Department of Veterinary Public Health and Food Safety, Unit of Food-borne Zoonoses and Veterinary Epidemiology, Istituto Superiore di Sanità in Rome. In relation to this work, the WHOCC has developed a subtyping PCR protocol for subtyping vtx1 and vtx2 genes. The protocol has now been blind tested on a reference collection of 63 VTEC strains in seven (two German, one Belgian, one Italian, two US and one Danish) independent laboratories with similar results, and will be included in the next VTEC EQA round for 2010 11. The results for detection of bfpa seem to indicate that both positive and negative reporting from two laboratories may not be trusted. One laboratory had 5/6 false positives and one laboratory failed to detect the one bfpapositive strain. Reporting on asta also varied from three to seven laboratories for the four asta-positive strains indicating that this gene may be underreported. Standardisation of both the nomenclature and methodology of bfpa, lpf, efa1/lifa and toxb genes needs further validation in order to be properly included in the EQA programme. 11

EQA for typing of VTEC TECHNICAL REPORT 6 Conclusions In conclusion, an average of 78% of the EU/EEA laboratories correctly O:H serotyped the ten VTEC strains representing 57% of the total number of reported serotypes to Enter-net in the period 2000 2006. Phenotypic performance was generally very high higher than 92% in the EU/EEA laboratories. Performance in gene detection was very high except for subtypes vtx1d and vtx2f. Six EU/EEA laboratories should review their genetic detection methodology for vtx2f and one laboratory should review their detection for vtx1d. Two EU/EEA laboratories should reduce their false-positive PCR results by looking into the possible reasons for this, including procedures for sample preparation of template DNA for PCR in order to minimise the risk of cross-contamination An average of 86% of the non-eu/sf laboratories correctly O:H serotyped the ten VTEC strains. Phenotypic performance was higher than 85% in the non-eu/sf laboratories. Performance in gene detection was very high except for subtypes vtx1d and vtx2f. Thirteen non-eu/sf laboratories should review their genetic detection methodology for vtx2f and seven laboratories should review their detection for vtx1d. Four non-eu/sf should reduce their false-positive PCR results by looking into the possible reasons for this including procedures for sample preparation of template DNA for PCR in order to minimise the risk of cross-contamination. Finally, one non-eu/sf laboratory should review their choice of primers for detection of the eae gene. 12

TECHNICAL REPORT EQA for typing of VTEC References 1. Bielaszewská M., Friedrich A W, Aldick T, Schurk-Bulgrin R, Karch H. Shiga toxin activatable by intestinal mucus in Escherichia coli isolated from humans: Predictor for a severe clinical outcome. Clin.Infect.Dis. 2006;43:1160-1167. 2. Friedrich A W, Bielaszewská M, Zhang W L, Pulz M, Kuczius T, Ammon A, and Karch H. Escherichia coli harboring Shiga toxin 2 gene variants: frequency and association with clinical symptoms. J.Infect.Dis. 2002;185:74-84. 3. Persson S, Olsen K E P, Ethelberg S, Scheutz F. Subtyping typing method for Escherichia coli Shiga toxin (Verocytotoxin) 2 variants and correlations to clinical manifestations. J Clin Microbiol. 2007;45:2020-2024. 4. Scheutz F, Ethelberg S. Nordic Meeting on detection and surveillance of VTEC infections in humans, p. 1-30. 2008;Statens serum institut. 5. de Sablet T, Bertin Y, Vareille M, Girardeau J P, Garrivier A, Gobert A P, Martin C. Differential expression of stx2 variants in Shiga toxin-producing Escherichia coli belonging to seropathotypes A and C. Microbiol. 2008;154:176-186. 6. Gobius K S, Higgs G M, a Desmarchelier P M. Presence of activatable Shiga toxin genotype (stx(2d)) in Shiga toxigenic Escherichia coli from livestock sources. J.Clin.Microbiol. 2003;41:3777-3783. 7. Jelacic J K, Damrow T, Chen G S, Jelacic S, Bielaszewská M, Ciol M, Carvalho H M, Melton-Celsa A R, O'Brien A D, Tarr P I. Shiga Toxin-Producing Escherichia coli in Montana: Bacterial Genotypes and Clinical Profiles. J.Infect.Dis. 2003;188:719-729. 13

EQA for typing of VTEC TECHNICAL REPORT Annex 1. Results of the participants by strain Table 4. Participants scores EU/EEA participants (percentage). The percentages represent the number of results in accordance with the intended result Strain No. O:H Serotype n = 13-16 O group n=15-19 (n=19 for O26, O103 & O157) H type n=13-16 (n=16 for H7 & H11) Sorbitol fermentation n=19 β- glucuronidase activity n=13 Haemolysin production n=9 VCA/ EIA n=8 eae n=21 vtx1 n=21 vtx2 ehxa n=21 n=2 0 Subtyping of vtx; Correct/tests AA1 71 73 75 100 100 100 100 100 95 100 100 4/7 (57%) BB2 94 94 88 84 100 100 100 100 100 95 100 3/3 (100%) CC3 79 88 80 100 100 100 100 100 95 71 100 10/11 (91%) DD4 57 63 87 95 69 89 100 100 100 100 95 3/8 (38%) EE5 79 100 73 100 69 100 100 95 100 95 100 2/6 (33%) FF6 79 88 80 100 100 100 100 100 95 100 90 4/8 (50%) GG7 93 100 87 100 92 89 100 100 100 95 100 3/3 (100%) HH8 88 100 82 100 92 100 100 100 100 95 100 3/3 (100%) ii9 38 44 57 95 100 100 100 100 91 100 100 4/5 (80%) JJ10 100 100 94 100 100 100 63 100 100 100 100 5/7 (71%) Average score 78 85 80 97 92 98 96 99.5 98 95 99 41/61 (67%) n: Number of laboratories participating per method Table 5. Participants scores non-eu/sf laboratories (percentage). The percentages represent the number of results in accordance with the intended result Strain No. Serotype n = 14-18 O group n=19-22 (n=21 for O26, O103, O111 and 22 for O157 H type n=14-18 (n=17 for H7) Sorbitol fermentation n=17 β- glucuronidase activity n=12 Haemolysin production n=11 VCA/ EIA n=9 eae n=20 vtx1 n=20 vtx2 n=20 ehxa n=16 Subtyping of vtx; correct/tests AA1 93 74 93 100 100 100 89 90 95 95 100 3/10 (30%) BB2 100 95 100 88 100 91 100 100 100 100 94 5/5 (100%) CC3 71 75 86 94 100 91 56 90 90 35 94 7/7 (100%) DD4 71 63 86 100 75 36 89 90 95 100 100 3/9 (33%) EE5 79 95 79 94 58 91 100 100 100 100 100 4/7 (57%) FF6 86 80 86 94 83 91 100 95 90 95 100 3/9 33%) GG7 93 90 93 94 75 73 100 95 100 100 100 4/4 (100%) HH8 89 100 89 100 83 91 100 95 100 100 100 4/4 (100%) ii9 79 63 86 94 100 91 89 95 65 95 94 3/4 (75%) JJ10 100 100 100 100 92 91 33 95 95 95 94 4/8 (50%) Average score 86 84 90 96 87 85 86 95 93 92 98 40/67 (60%) n: Number of laboratories participating per method 14

TECHNICAL REPORT EQA for typing of VTEC Table 6. Results of vtx1 and vtx2 subtyping EU/EEA and non-eu/sf laboratories. Numbers in brackets correspond to the number of laboratories reporting this result Strain No. Subtype by the WHOCC a EU/EEA participants results b Non-EU/SF participants results Number of laboratories getting the correct vtx subtypes/ number of laboratories testing Subtype Variant Subtype EU/EEA Non-EU/SF AA1 vtx1c + vtx2b vtx1c + vtx2b (4) vtx1c + vtx2b (3) 4/7 3/10 vtx2b (2) vtx2d (1) vtx2d (3) vtx2cd-neg; vtx2e-neg (1) vtx2e-absent (1) vtx1 + vtx2 (1) vtx1 (1) BB2 vtx1a vtx1a (3) vtx1a (5) 3/3 5/5 CC3 vtx2f vtx2f (10) vtx2g (1) vtx2f 10/11 7/7 DD4 vtx2c + vtx2d vtx2c + vtx2d (2) vtx2a (2) vtx2d (1) vtx2-nv206 (1) vtx2c (1) EE5 vtx1a + vtx2a vtx1a + vtx2a (2) vtx1a (1) vtx2a (2) vtx2c + vtx2d-o174- EC1720a (1) vtx2c + vtx2d (3) vtx2a + vtx2c (2) vtx2a + vtx2b (1) vtx2c (1) vtx1 + vtx2 (1) vtx2a (1) vtx2a-o157-sf-258-98 (1) vtx1a + vtx2a (4) vtx2cd-pos; vtx2e-neg (1) vtx2a + vtx2b (1) vtx2 untypable (1) 3/8 3/9 3/6 4/7 FF6 vtx1c + vtx2b vtx1c + vtx2b (4) vtx2a (1) vtx2b (2) vtx2d (1) vtx1c + vtx2b (3) vtx2d (2) vtx1 + vtx2 (1) vtx2cd-neg; vtx2e-neg (1) vtx1a + B subunit (1) vtx2e-absent (1) 4/8 3/9 GG7 vtx1a vtx1a vtx1a 3/3 4/4 HH8 vtx1a vtx1a vtx1a 3/3 4/4 ii9 vtx1d vtx1d (4) vtx1a (1) vtx1d (3) vtx1a (1) 4/5 ¾ JJ10 vtx2c vtx2c (5) vtx2a (1) vtx2b (1) vtx2c (5) vtx2a (1) vtx2a + vtx2c (1) vtx2a + vtx2b O157 (1) 5/7 5/8 a Results for vtx1 were by a triplex PCR (unpublished) with specific primers for vtx1a, vtx1c and vtx1d, and for vtx2 by a prototype PCR (unpublished) with specific primers for the seven subtypes: vtx2a, vtx2b, vtx2c, vtx2d, vtx2e, vtx2f and vtx2g. b The submitted results for subtyping of vtx genes were interpreted as follows: stx as vtx; vtx1 as vtx1a; vtx2 as vtx2a; vtx2d- Ount and vtx2-o118 as vtx2b, vtx2v-ha as vtx2c and vtx2-nv206 as vtx2d. 15

EQA for typing of VTEC TECHNICAL REPORT Table 7. Results of additional virulence genes by all participating laboratories Strain No Antigen related genes Virulence genes (n) A1 flich8 eaaa espi lpfa, lpfa2-1 iss iha (3) BB2 flich7 asta (3) bfpa a eae γ (3), eae γ1 efa1 (4) espa_o157h11 espf espj etpd (4) iha iss lpfo157/oi-154, lpfo157/oi-141, lpfo141, lpfa1-3, lpfa2-2 nlea nleb nlec pagc tccp tir_o157h7 CC3 asta (7) bfpa (2) b cba cdt cif cma eae β (3), eae β1 espb espa_o103h2, espf_o103h2 iss lpfa1-2, lpfao26nlea nlec tccp tir_o103h2 DD4 cif eae β (3), eae β1 espa_o103h2, espf_o103h2 espb espj espi espp (3) iha (4) katp (5) lpfa2-1, lpfa nlea nleb nlec tere tccp tir_o103h2 EE5 rfbo111 flich8 bfpa a cba celb cif eae θ (2) efa1 (4) espa espf espj iha (3) lpfa, lpfa1-2, lpfa2-1, lpfao26 nlea nleb nlec pagc tccp tir tere wbdl FF6 rfbo113 asta (6) cba celb cma eaaa espi iha (4) saa senb 16

TECHNICAL REPORT EQA for typing of VTEC Strain No Antigen related genes Virulence genes (n) GG7 wzxo103 bfpa a cba cma cif eae ε (3) efa1 (4) espa_o49h12 espb espf_o103h2 espj etpd (4) katp (5) lpfa1-2, lpfao26 nlea nleb pagc tccp tir_o103h2 HH8 wbuao26 bfpa a eae β (3) eae β1 efa1 (3) iha (2) lpfa1-2, lpfa2-1, lpfao26 tere ii9 cba cma mchb mchc mchf lpfa1-5, lpfo141, lpfo157/oi-154, lpfo157/oi-141 JJ10 rfbo157 (2) flich7 asta (2) bfpa eae γ (3), eae γ1 efa1 (2), efa1-1 espa_o157h11 espf espj espp (3) etpd (5) iha (4) katp (5) lpfo157/oi-154, lpfo157/oi-141, lpfa1-3, lpfa2-2, lpfo141 nlea nleb nlec pagc rfbe tccp tere tir_o157h7 toxb (2) uida a One laboratory has these six strains positive for bfpa. These strains were found negative by one participant and the Centre except for strain CC3. b Found negative by two participants for bfpa and by one participant for bfpb. Genes found by participating laboratories asta, bfpa, cba, cdt, celb, cif, cma, eae α, eae β, eae β1, eae γ, eae γ1, eae ε, eae θ, efa1, efa1-1, espa, espa_o103h2, espf_o103h2, espa_o157h11, espa_o49h12, espb, espf, espf_o103h2, espi, espj, espp, etpd, eaaa, iha, iss, katp, lpfa, lpfa1-2, lpfa2-1, lpfao26, lpfa1-5, lpfo141, lpfo157/oi-154, lpfo157/oi-141, lpfa1-3, lpfa2-2, mchb, mchc, mchf, nlea, nleb, nlec, pagc, rfbe, senb, saa, tccp, tere, tir, tir_o103h2, tir_o157h7, toxb, uida and wbdl. 17

EQA for typing of VTEC TECHNICAL REPORT Annex 2. List of the participants ECDC-funded EU/EEA participants Laboratory Institution Country Bereich Humanmedizin, Institut für Medizinische Mikrobiologie und Österreichische Agentur für Gesundheit und Austria Hygiene Bereich Humanmedizin Ernährungssicherheit GmbH Dept. Microbiology Universitair Ziekenhuis Belgium Nat Ref Lab for Enteric Pathogens National Centre of Infectious and Parasitic Diseases Bulgaria Microbiology Department Nicosia General Hospital Cyprus NRL for E. coli and Shigella National Institute of Public Health Czech Republic Foodborne bacteria and typing, Dept. of Microbiological Surveillance Statens Serum Institut Denmark and Research Head ESYV Reference Unit, Laboratory of Enteric Pathogens Specialist and Reference Microbiology Division, Health England Protection agency Unit of Bacteriology National Institute for Health and Welfare (THL) Finland Centre National de Référence des Escherichia coli et Shigella Unité Biodiversité des Bactéries Pathogènes Emergentes, France Institut Pasteur NRC Salmonella and other Enterics Robert Koch Institute, Bereich Wernigerode Germany Department of Microbiology National School of Public Health Greece B. Johan National Center for Epidemiology Hungary SWAHB, Public Health Laboratory Cherry Orchard Hospital Ireland Dipartimento di Sanità Alimentare e Animale Istituto Superiore di Sanità Italy Microbiology department National Public Health Surveillance Laboratory Lithuania Laboratoire National de Santé Division Microbiologie Luxembourg Diagn. Lab. for Infectious Diseases and Perinatal screening National Institute of Public Health and the Environment, Centre Netherlands for Infectious Dept. of Foodborne Infections, Division of Infectious Disease Control Norwegian Institute of Public Health Norway Division of Infectious Disease Control Department of Bacteriology National Institute of Hygiene; Narodowy Instytut Zdrowia Poland Publicznego Reference Laboratory for Molecular Epidemiology National Institute of Research-Development for Microbiology Romania and immunology Cantacuzino Laboratory of Enteropathogenic bacteria, Department of Medical National Institute of Public Health Slovenia Microbiology Sección de Enterobacterias, Servicio de Bacteriología Carretera de Majadahonda a Pozuelo, Centro Nacional de Spain Microbiologia Dept. of Bacteriology Swedish Institute of Infectious Disease Control Sweden ECDC-funded non-eu/eea country Laboratory Institution Country National Reference Laboratory for Enteric Pathogens, Refik Saydam Public Health Agency Turkey Communicable Diseases Research Department Self-funded participants Laboratory Institution Country Servicio Fisiopatogenia Inei-Anlis Dr. Carlos G. Malbrán Argentina Microbiological Diagnostic Unit Public Health Laboratory Department of Microbiology and Immunology, University of Australia Melbourne Enteric and Food Microbiology Laboratory ICDDR,B Bangladesh Secao de Bacteriologia, Laboratorio Central De Saude Publica Instituto Adolfo Lutz Brazil Head Identification and Serotyping, Enteric Disease Program National Microbiology Laboratory Canada E. coli Laboratory, Laboratory for Foodborne Zoonoses Public Health Agency of Canada Canada Reference Laboratory for Escherichia coli, GREMIP, Faculté de Université de Montréal Canada médecine vétérinaire Service de microbiologie Laboratoire associé au CNR E. coli- Hôpital Robert Debré France Shigella Abteilung Mikrobiologischer Verbraucherschutz Institut für Hygiene und Umwelt Germany Nationales Referenzlabor für Escherichia coli (NRL-E. coli) Bundesinstitut für Risikobewertung (BfR) Germany Department of Bacteriology National Institute of Cholera and Enteric Diseases (Indian India Council of Medical Research) Department of Bacteriology National Institute of Infectious Diseases Japan 18

TECHNICAL REPORT EQA for typing of VTEC Public Health Laboratory Department of Public Health, Faculty of Medicine, National Mexico Autonomous University Mexico Enteric Reference Laboratory; ESR NCBID Wallaceville Institute for Environmental Science and Research Ltd New Zealand Microbiology Department Research Institute for Tropical Medicine, Department of Philippines Health Enteric Diseases Reference Unit (EDRU) National Institute for Communicable Diseases, NHLS South Africa (previously SAIMR) Nationales Zentrum für enteropathogene Bakterien (NENT) Institut für Lebenmittelsicherheit und -hygiene Switzerland E. coli Reference Center The Pennsylvania State University USA National Reference Laboratory for Escherichia coli and Shigella, Centers for Disease Control and Prevention USA DASH, Unir 7 U.S. FDA, HFS-711, Washington USA The National Institute For Food Control Vietnam 19